“Organizations managing AI as isolated experiments will lose to those governing it as a portfolio.”
— Taopheek Babayeju, CEO, iCentra
The Project Management Office has spent the last decade redefining its own purpose. The most forward-looking PMOs moved from project administration to portfolio governance — from tracking timelines and budgets to actively managing strategic alignment, resource trade-offs, and value realization across a portfolio of organizational investments. This transition — from PMO to Value Management Office — represented a genuine maturation of the function: from delivery mechanism to governance institution.
That maturation now matters more than ever. Because the next major challenge in enterprise investment governance is not traditional projects. It is AI.
The AI portfolio problem
Most organizations experiencing AI momentum right now are not running one AI initiative. They are running many — often across multiple functions, with multiple vendors, at different stages of maturity, with different sponsors and different definitions of success. Marketing has a generative AI content tool. Operations has a process automation pilot. Finance is running an AI-assisted forecasting model. HR is piloting an AI screening tool. Customer service has a chatbot. Each initiative has a business case. Few have been evaluated against each other.
This is not AI adoption. This is AI fragmentation.
The consequences are predictable and already visible in many organizations. Resources allocated to AI initiatives compete without coordination, creating bottlenecks that individual sponsors cannot resolve. Data governance obligations accumulate across disconnected initiatives without centralized oversight. Risk exposures multiply without anyone holding a portfolio view of aggregate AI risk. And return expectations — set in isolation, by initiative — are rarely tracked against a coherent portfolio performance standard.
The organizations managing AI as a collection of isolated experiments will eventually be surpassed by those governing it as a coherent portfolio. The question is not whether portfolio governance is needed. The question is who provides it.
Why the PMO is the natural answer
The Value Management Office — the evolved PMO — exists to do exactly what AI investment governance requires. It holds portfolio visibility across all active investments. It manages resource allocation at the portfolio level, resolving conflicts that individual initiative sponsors cannot see. It defines and monitors performance standards that connect investment activity to strategic return. It provides the governance architecture within which individual initiatives operate, without displacing ownership of those initiatives from their business sponsors.
These are not new capabilities. They are the capabilities that mature PMOs and VMOs have been building for years. AI governance does not require a new institution. It requires an existing institution to extend its remit.
There are three ways the PMO/VMO can immediately extend its governance role to cover AI investment.
Portfolio registration and classification
The starting point for AI portfolio governance is visibility. Before any AI initiative can be governed effectively, it must be registered — identified, scoped, classified, and connected to the portfolio. Registration should capture the strategic priority the initiative serves, resources committed (budget, data, technology, people), expected return and the mechanism for measuring it, the risk profile (data governance obligations, algorithmic decision risk, vendor dependency), and the accountability assignment — who owns the investment on behalf of the organization, not just who manages the project.
Many organizations will find, when they conduct this exercise, that they cannot accurately identify all active AI initiatives across the enterprise. That discovery is itself important governance information. An organization that cannot inventory its own AI investments cannot govern them.
Strategic alignment and resource governance
Portfolio governance requires the ability to make trade-off decisions: to redirect resources from a lower-priority investment to a higher-priority one when constraints arise, to defer an initiative that no longer fits strategic priorities, and to scale an investment that is outperforming expectations.
Applying this discipline to AI investments requires two things. First, AI investments must be evaluated against the same strategic priority framework as all other organizational investments — not as a separate category that bypasses the portfolio governance process. Second, the VMO must have the mandate and executive sponsorship to make resource governance decisions across AI initiatives, not merely to track them.
Without this, the PMO remains a scorekeeper rather than a governor. It can tell leadership that the AI portfolio has a resource conflict. It cannot resolve it. The institutions that make portfolio governance effective are those in which the governance function has authority, not just visibility.
Benefits realization beyond project closure
The governance gap in most AI initiatives is not at initiation. It is after deployment. Organizations invest in AI tools, deploy them, and move on to the next initiative — without a mechanism for systematically tracking whether the deployed AI is actually producing the returns it was approved to deliver. Were efficiency gains realized? Were cost reductions captured? Were decision outcomes improved?
Benefits realization governance for AI investments requires defining, at initiation, what success looks like in measurable terms — and then maintaining accountability for those measures through deployment and into the operational phase. This is standard VMO discipline applied to a new class of investment. It is also one of the most frequently missing elements of enterprise AI governance.
The leadership conversation that needs to happen
PMO and VMO leaders who want to extend their governance mandate to cover AI investments need to have a specific conversation with executive leadership. That conversation is not primarily about AI. It is about portfolio discipline.
The argument is straightforward: the organization is making significant AI investments. Those investments need portfolio-level governance — visibility, resource management, performance accountability, and risk oversight. The PMO/VMO is the institutional mechanism for exactly this. Extending its mandate to cover AI investments is not expansion for its own sake. It is the application of existing governance capability to a new and growing category of organizational investment.
That argument, made clearly and connected to current AI portfolio fragmentation, is compelling to most executive leadership teams — particularly those beginning to recognize that their AI investments are producing more activity than measurable value.
What this means for the future of the PMO
The PMOs that establish themselves as the governance anchor for AI investment over the next 12 to 24 months will be significantly more strategically important than those that remain in a traditional project administration role. The governance challenge posed by enterprise AI is one of the most significant portfolio management challenges of the current decade. The institutions that step into that challenge with the right capabilities will become indispensable to their organizations’ strategic decision-making.
This is not a technical evolution for the PMO. It is a governance one. And it is precisely the evolution that the profession has been building toward.
Organizations managing AI as isolated experiments will lose to those governing it as a portfolio. The PMO — evolved into its full potential as a Value Management Office — is the institution that makes portfolio governance possible. The organizations that recognize this early will be the ones with both the AI investments and the governance infrastructure to make those investments count.
Taopheek Babayeju is the CEO of iCentra, a global technology and business solutions company specializing in enterprise governance, transformation, and execution capability. Learn more at icentra.com.
